Data are created by sensors and aggregated by the company (institution) that has an interest in those data (and probably owns/deployed the sensors). Data creation, as noted, is a fall out of the Digital Transformation, companies digitise their environment and create data that are used by them and can be leveraged through sharing.
In general, however, the real value for third parties does not rise from the access to data, rather from correlation of several streams of data, in other words, data relationship is often more valuable than the data themselves. This is the area of context and meaning:
- by correlating data it is possible to evaluate them in a specific context (like a temperature data can assume different meaning depending on the context: if it is summer you can expect the home temperature to be higher than in winter, however if you know that a fireplace has been lit it is obvious to expect that the surrounding area will show an higher temperature than normal; a heartbeat around 100 may raise an alarm, but if occurs in a person that is running there is nothing to be worried about);
- by analysing data in a given context and correlating them to other data streams one could extract meaning. A traffic jam detected by a road safety camera can become meaningful when correlated to historical data showing that every Friday at that time a traffic jam occurs. On the other hand if historical data shows that seldom there is a traffic jam then further correlation may detect a flooding with road block in another part of the city that is diverting traffic thus creating a traffic jam.
Correlation and analytics create new data, usually called “meta-data”. It may be difficult for an end user to access different streams of data, correlate and analyse them to generate meta data. This is where third parties can step in and offer meta-data as services. Sometimes, these meta-data can result from an end user request through a service API, like, in the previous example, a request of “what is the fastest route to get to point B?”. Answering this question requires a visibility on current road situation and the ability to forecast the evolution over time. This in turns requires an understanding of what is going on, something that needs context and meaning.
Contextualisation and meaning can be obtained through AI technologies that can also be based on historical, statistical data. This requires historical records and again it may be a business space for specialised companies. These companies can also develop statistical analytics and sell them as a service.
Contextualisation can be referred both to the interpretation of data (that is what I have been discussion before) but it can also be referred to the user context, adapting the response, and interpretation of a query, to the user context. In this area Personal Digital Twins will be playing a growing role in the future, also for privacy reasons (people may be reluctant to share information on their context, yet they still would need to get information that is relevant in their context).
Finally, data correlation and meta-data creation can lead to the creation of an ambient in the cyberspace supporting simulation and digital operation. This, again, is an area where Digital Twins can be used. Industry 4.0 is going to make use of digital ambients created through data correlation and analytics.